HTMOTW (Home Town Memo Of The Week - for the Seattle Metro Area)
Date: Thu, 6 Jan 94 13:45:36 PST
Subject: HTMOTW (Home Town Memo Of The Week - for the Seattle Metro Area)
[I have a few questions about this memo:
(+) is it for real?
(+) what is "mindshare"?
(+) is this the way marketing is usually done?
(+) wasn't "mission-critical" last year's buzzword?
(+) is verbifying "architect" worse that coining "marketecture"?
(+) is this memo really, really for real?
I guess I've just led a sheltered life, working for film moguls, large Wall
Street law firms and The Phone Company, where everybody is so nicey-nice... -psl]
From: Keith Bostic <vangogh.CS.Berkeley.EDU!bostic>
Send-To: uunet!microsoft.com!steveb uunet!microsoft.com!jonl
During the past few months, I have grown increasingly
uncomfortable with our collective response to the threat of Notes.
While the positioning of this product was essentially botched by
Lotus during the first several years of its existence, it has noneth
eless gained tremendous significance in the eyes of both the industry
and our major customers. While it remains the situation that few can
tell you what Notes actually is, customers now spontaneously think of
Notes when the words "workgroup" or "groupware" are mentioned. What
comes to *your* mind when "groupware" is mentioned?
When Notes had less visibility and was being ineffectively
marketed, it was much easier for us to dismiss the situation as a
temporary phenomenon that we could effectively combat in any of a
number of ways, from either our systems or applications efforts.
However, Lotus has now gained the moral high ground in the industry
with Notes, as clearly demonstrated by the rise of their stock over
the course of the year. In the eyes of the press and analysts, this
product can do no wrong, even with all of its obvious weaknesses.
The situation now appears to be out of our control; as you
well know, virtually every one of our major corporate accounts has
cornered one of you and asked how our applications will work with
Notes. Due to clever marketing, Lotus has positioned Notes in these
customers' minds as a "platform", convincing them that it has some
kind of underlying architecture that resembles system software.
Although we've estimated that Lotus has only sold about 350,000
copies worldwide, the high acceptance rate of Notes within our major
accounts is extremely disturbing. These are the only people who care
about systems architecture; if we fail to capture their mindshare, do
you have any idea how expensive it will be to regain it?
Between Lotus' mindshare and Novell's network dominance, I am
concerned that our enterprise strategy message is being reduced to
irrelevance in many accounts. Our success to date has been a result
of selling desktop operating systems and desktop applications, but I
must reinforce that moving forward, whoever owns the network
infrastructure will increasingly own the desktop applications.
Customers are increasingly looking toward single-vendor solutions.
Since Novell doesn't have applications, this means that Notes could
lead to increased sales of cc:Mail and other Lotus products. We MUST
control the infrastructure moving forward or we will fail to give our
applications the leverage that they need for continued success!
But as we learned when launching WFW, it has proven
surprisingly dangerous for us to publicly assail Notes.
Notes-bashing is not a sensible solution, first because we can't risk
alienating our major accounts who have already embraced the product,
and second because we don't currently provide an acceptable
alternative. We must take an indirect approach to the problem,
positioning it as an isolated application and as complementary to our
products. The "one-two punch" of our Anti-Notes message must be 1) t
o disarm the situation by complimenting Notes on the desktop,
demonstrating how well the product integrates with Microsoft desktop
applications through standard Microsoft technologies such as OLE and
MAPI, and 2) to introduce uncertainty in the customer by making it
very clear that our systems infrastructure is much more robust than
Notes will ever be, that they don't have any experience with systems
or infrastructure, and that there will be no need for their back-end
once our systems are in place.
In short, make the customer feel OK that he made both the
Notes purchase decision and the Microsoft Office purchase decision,
but make him think very carefully about enterprise-wide deployment.
If, by these tactics, we can delay enterprise-wide adoption of Notes
in our major accounts by a measurable degree, the probabilities will
be high that the customer will then force Lotus to rework Notes to
work on top of our infrastructure. Thus, while Notes will still
exist on the desktop, we will have won the enterprise computing
infrastructure. And once Notes is contained on the desktop, it will
fail on its own due to its inherent lack of top-end programming
capabilities, as compared to our own desktop strategy.
In summary, here are the 10 Anti-Notes Commandments that you
and your people should memorize:
I Notes is a good Windows product, and a great OLE application
that works exceedingly well with all applications in the Microsoft
Office suite. But be careful when using it with Lotus OLE-enabled
products, because we're told that they require "back doors" around
OLE in order to be useful when used with Notes.
II As an application development environment, Notes is
extremely weak. It lacks a true programming language, is not
transscriptable, and you have to drop into "C" in order to do
III Because of its 16-bit heritage, Notes is rife with 64K
limits that show up to the user in very inconvenient ways. The
32-bit version of Notes is simply a recompiled 16-bit version, and
still has all of these limitations!
IV Notes isn't architected for the enterprise. Dating back to
1984, it was designed for little workgroups to be connected by phone
lines, not for an enterprise with tens of thousands of users.
V Notes duplicates many of the functions that are in the
operating system, such as Mail, Directory and Security. Since the
weak point of any secure system is in its management by humans,
incompatible duplicate security subsystems and directories result not
only in additional management cost, but also weakened security.
VI Lotus as a company doesn't know how to build or support
large-scale systems. The back-end of a distributed object store
should be built and supported by a company that already supports tens
of millions of operating system users, not by a company whose only
claim to fame is a spreadsheet.
VII Lotus sells two messaging products, cc:Mail and Notes,
that work in completely different and incompatible manners. While
they have finally seemed to have learned the art of "marketecture"
with their LCS announcement, slides should not be confused with a
well-architected product. MAPI and the other WOSA service provider
interfaces already have the support of hundreds of third-party
developers, and will work flawlessly with EMS and all of Microsoft's
future operating systems.
VIII The Notes Server has proven to be a pitiful performer on
OS/2, but we've heard that it runs extraordinarily well on
Windows/NT. Because Notes has no built-in support for managing large
numbers of servers, we'd suggest that if you plan on deploying Notes
servers, you use Windows/NT in conjunction with Hermes in order to do
distributed systems management.
IX Cairo has a distributed, replicated object store that is
much more robust, efficient, and scaleable than Notes, and in
addition it's an open system having been built on DCE RPC. DEC is
working with us to ensure that our distributed object technology is
portable to other vendors' environments.
X The fact that Lotus is working on many incompatible code
streams for Notes on multiple platforms is stretching Lotus past its
limits. It has already caused them to miss schedule milestones a
number of times, and thus be an unreliable supplier of software.
Notes' support of multiple platforms is not unique - all of
Microsoft's applications are programmed to the Win/32 API, and will
port to Unix, Macintosh, and OS/2 with a simple recompile.
Finally, here's a quick-reference guide that can be used to
make potential clients nervous or insecure about Notes, while not
alienating them in case they actually do buy Notes:
1. Remember that, in the presence of a client, you should
always "be apparently respectful" to Notes, especially if that client
is a major account.
2. Always focus on how well Microsoft Office applications work
with Notes because Notes was an OLE pioneer, working very closely
with Microsoft. Always add that the other Lotus applications were
"followers", and have always had abysmal OLE support.
3. Use a number of methods to introduce uncertainty about the
"back end" in the customer's mind. The following have been tested
and are known to work:
a. They have a directory, but I've heard that it doesn't scale
to the size of your organization, and the one in Notes Version 3 was
rewritten and has problems.
b. Due to the large Microsoft development community, we
already know that virtually every vendor of messaging programs,
systems and services are working on Microsoft EMS integration. I'm
really not sure yet how Lotus is going to integrate with EMS.
c. The Notes security stuff was good when it was developed,
but that was in 1984. And I've heard that they've had some security
problems lately - related to the fact that it wasn't in the operating
d. The biggest problems people seem to have with Notes is in
getting support from Lotus. The integrators who install Notes seem
incapable of dealing with any but the simplest of problems, and Lotus
itself is very slow at problem resolution - they clearly don't
understand what it takes to run a mission-critical enterprise system!
e. The next version of NT, Cairo, has a fully *replicated*,
distributed object file system, and I'm not yet sure how Lotus is
going to integrate with it. A very large development community is
already working with the prerelease development kits. You really
should ask Lotus about it.
4. If you're sure that they're going to install Notes (or
already have bought it), make sure that you pitch NT as the preferred
server environment. Once NT is "in the door", it will be much easier
for us to supplant Notes on the back-end with NT Version 2 (Cairo).
"I'm sure that you've heard that the Notes Server is a dog on
OS/2, but they've been working with NT for years and it's a screamer,
especially due to the SMP support."
"Yes, I've heard that the NLM is faster, but it's VERY
unstable and can't scale due to the lack of SMP."
"Because Notes is difficult to manage remotely, you can use
Hermes to manage your Notes Servers."
Finally, lest you feel that we're only dealing with Notes
with words, let me remind you of our current technical efforts, on
both the systems side as well as in applications:
- EMS will be a combination back-end message router and
distributed client/server message store. It will have both public
and private relational "message databases", similar to Notes forums,
that can be fully replicated with other EMS systems. Message da
tabases are customizable, and front-end tools are included to permit
the user to construct "conferencing" and other shared message store
applications. It has a tremendously impressive message store,
directory, and routing administration tool known as "Mailbeat",
public key security, and native X.400 and X.500 support. We will be
bundling EMS with every copy of NT Advanced Server and Cairo on a
permanent basis starting at first shipment of EMS. You should think
of EMS as the "back-end companion" to Chicago.
- Chicago will finally eliminate any reason to buy a
Macintosh. It will have an impressive object-oriented UI, integrated
networking, integrated VB, integrated rich text editor, integrated
Mail and Messaging, full MAPI 1 support with integrated service pr
oviders for out-of-the-box communications with CompuServe, EasyLink,
America OnLine, and MicroServe, our new on-line service. Chicago
will make extensive use of OLE2 in the "shell", creating the
appearance of a seamless system. Chicago, combined with integrated
mail, VB, and rich text editor, with NT and EMS on the back-end,
should give nobody any more *practical* reason to purchase Notes in
order to build distributed workgroup applications.
- Windows NT Version 2, or Cairo, is a completely
object-oriented back-end. The user-perceivable benefit is to extend
the notion of a fully replicated message store back-end, achieved in
EMS, to the entire file system. Wherein Chicago was the perfect OLE2
front-end, Cairo will be the perfect OLE2 back-end.
- VBA is being enhanced to have full VB functionality, and
will be present in all Office applications. Included in this
functionality is a set of EMS "controls" that include not only
message store and directory traversal, but also routing and
- VBA is being enhanced so that VBA applications themselves
are stored inside of "standard containers". As a result, VBA
applications can be fully distributed using either the EMS or Cairo
replicated message and object stores.
- Office will continue to provide the best OLE2 container
support in the industry.
- We will be producing a "Chicago Office" suite to counter any
potential Notes-based suite threat. After installation, all Office
applications will have apparent "shell-level" integration.
Good luck, and please keep me informed as to your progress.
© 1994 Peter Langston